Traffic Flow Estimation
Overview
Traffic flow estimation reconstructs current traffic conditions and predicts how vehicles move through roads, intersections, and urban networks. It draws on fixed sensors, vehicle trajectories, cameras, radar, mobile-network data, and aggregated counts, often combining sources when coverage is sparse. Common approaches include statistical and Bayesian models, traffic-flow physics and conservation constraints, graph neural networks and other spatiotemporal models, and optimization or reinforcement learning for origin–destination demand. Methods can perform well on road-sensor and trajectory benchmarks, but reliable transfer across cities and changing conditions remains difficult, as does balancing predictive accuracy with physical consistency, uncertainty estimates, and computational cost.
Latest papers
51 papers3 papers in the last four weeks, down 62% on the four weeks before. 0.0% of all new papers.